The strength of the association between prognostic variables, and the outcome of interest was expressed as a hazard ratio and corresponding 95% confidence interval (CI) calculated [ 32, 33].
Seven studies [ 22, 24- 29] omitted details of random variability and measures of association between prognostic variables and outcome (or differences between prognostic groups).
Cox proportional multivariate analysis of relationships between prognostic variables and survival are shown in Table 4.
We performed a concordance analysis using the χ2 test between prognostic variables and the presence or absence of polysomy 17 to detect any differences between polysomic and non-polysomic subgroups.
Relationships between prognostic variables are difficult to interpret, and better means of prognostication are needed since there are currently no reliable methods for estimating survival in this patient population.
Statistical independence between prognostic variables was evaluated by multivariate analysis using the Cox proportional hazard model.
Univariate correlations between prognostic variables and survival outcomes were carried out using the Kaplan-Meier method.
However, the sample size was sufficiently large to demonstrate a statistical correlation between the classical prognostic variables of tumour size, lymph node status, lympho-vascular invasion, metastases and the NPI.
A Cox proportional hazards model was fitted to investigate trends between biomarker expression, the standard prognostic variables of Gleason grade, stage and survival.
